Glutathione Is Involved in the Granular Storage of Dopamine in Rat PC12 Pheochromocytoma Cells: Implications for the Pathogenesis of Parkinson’S Disease

Home , Glutathione disulfide, PC12 cell line

The Journal of Neuroscience, October 1, 1996, 16(19):6038–6045

Glutathione Is Involved in the Granular Storage of Dopamine in Rat PC12 Pheochromocytoma Cells: Implications for the Pathogenesis of Parkinson’s Disease

Benjamin Drukarch, Cornelis A. M. Jongenelen, Erik Schepens, Cornelis H. Langeveld, and Johannes C. Stoof Department of Neurology, Graduate School Neurosciences Amsterdam, Research Institute Neurosciences Vrije Universiteit, 1081 BT Amsterdam, The Netherlands

Parkinson’s disease (PD) is characterized by degeneration of of DA stores with the tyrosine hydroxylase inhibitor ␣-methyl- dopamine (DA)-containing nigro-striatal neurons. Loss of the p-tyrosine. In the presence of ␣-methyl-p-tyrosine, refilling of antioxidant glutathione (GSH) has been implicated in the patho- the DA stores by exogenous DA reduced GSH content back to genesis of PD. Previously, we showed that the oxidant hydro- control level. Lowering of PC12 GSH content, via blockade of gen peroxide inhibits vesicular uptake of DA in nigro-striatal its synthesis with buthionine sulfoximine, however, led to a neurons. Hydrogen peroxide is scavenged by GSH and, there- significantly decreased accumulation of exogenous [3H]DA fore, we investigated a possible link between the process of without affecting uptake of the acetylcholine precursor vesicular storage of DA and GSH metabolism. For this purpose, [14C]choline. These data suggest that GSH is involved in the we used rat pheochromocytoma-derived PC12 cells, a model granular storage of DA in PC12 cells and that, considering the system applied extensively for studying monoamine storage molecular characteristics of the granular transport system, it is mechanisms. We show that depletion of endogenous DA stores likely that GSH is used to protect susceptible parts of this with reserpine was accompanied in PC12 cells by a long- system against (possibly DA-induced) oxidative damage. lasting, significant increase in GSH content the extent of which Key words: Parkinson’s disease; oxidative stress; glutathione; appeared to be inversely related to the rate of GSH synthesis. dopamine; PC12 cells; neurotransmitter uptake; vesicular A similar increase in GSH content was observed after depletion storage

Parkinson’s disease (PD) is characterized primarily by a loss of major antioxidant in the brain. Thus, GSH has been shown to dopamine (DA) in the striatum caused by degeneration of DAer- detoxify (organic and inorganic) hydroperoxides including the gic neurons in the zona compacta of the substantia nigra (SN) H2O2 formed during the oxidative metabolism of DA (Spina and (Gibb and Lees, 1991). Although the cause of PD is still unknown, Cohen, 1988, 1989). oxidative stress has been implicated as a pathogenetic factor. To prevent breakdown of DA before its release from presyn- According to this so called “free radical hypothesis,” the degen- aptic endings, the transmitter is taken up and stored in presynap- eration of the nigro-striatal system in PD is related to the rela- tic granules. This process is effected through the action of a tively high exposure of these neurons to reactive oxygen species transporter protein (Johnson, 1988). Interference with granular

(ROS), in particular hydrogen peroxide (H2O2), produced during storage, for instance via blockade of the ligand-binding site on the both the enzymatic (monoamine oxidase-catalyzed) and nonenzy- transporter, leads to an enhanced turnover of DA, inducing con- matic (auto-oxidative) breakdown of DA (Adams and Odunze, siderable oxidative stress as a consequence of H2O2 production 1991; Olanow, 1992). Thus, DAergic cell death in PD may be (Spina and Cohen, 1989). Recently, we have shown that, apart caused by an overproduction of ROS and/or a diminished protec- from generating oxidative stress, H2O2 is able to inhibit storage of tion against them. Evidence to support the “free radical hypoth- DA in the nigro-striatal system through an as yet undefined esis” has come from postmortem investigations of PD brains, interaction with the granular uptake mechanism (Langeveld et al., which consistently show an increase in the indices of oxidative 1995). Interestingly, the DA storage system appeared to be sub- stress in the SN at the time of death (Hirsch et al., 1991; Jenner, stantially more sensitive to H2O2 compared with that of other 1993). One of these indices, possibly even preceding the loss of transmitters, such as noradrenaline (NA) and acetylcholine DA (Dexter et al., 1994), is a decrease in the level of glutathione (ACh) (Langeveld et al., 1995). Based on these data, we proposed (GSH) in the Parkinsonian SN (Riederer et al., 1989; Sofic et al., that the degeneration of the nigro-striatal system in PD may be 1992; Jenner, 1993). GSH, the most abundant free thiol in mam- the outcome of a self-sustaining cycle that is initiated by lowering malian cells (Meister and Andersson, 1983), is considered to be a of the capacity to scavenge the H2O2 released during normal DA metabolism, leading to a decrease in the granular storage and Received March 19, 1996; revised July 12, 1996; accepted July 18, 1996. enhanced breakdown of DA, inducing more oxidative stress, dis- We thank Dr. Robbert J. Slingerland (University of Amsterdam) for providing us turbance of neuronal function and, eventually, cell death. Consid- with the PC12 cell line. ering our observation that H2O2 preferentially interferes with Correspondence should be addressed to Dr. Benjamin Drukarch, Research Insti- presynaptic storage of DA and the fact that, in the mammalian tute Neurosciences Vrije Universiteit, Department of Neurology, Boechorststraat 7, 1081 BT Amsterdam, The Netherlands. brain, H2O2 is detoxified mainly by GSH (Maker et al., 1981; Copyright ᭧ 1996 Society for Neuroscience 0270-6474/96/166038-08$05.00/0 DiMonte et al., 1992), whose levels are decreased in PD, we Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine J. Neurosci., October 1, 1996, 16(19):6038–6045 6039 initiated a series of in vitro experiments to investigate a possible previously (Westerink and Mulder, 1981; Van Muiswinkel et al., 1995), linkage between granular uptake of DA and GSH metabolism. using a Guardcolumn (Phase Sep, Deeside, UK) in combination with a reverse-phase column filled with Spherisorb (3 ␮m, SRODS2; Phase Sep) For this purpose, we used a rat pheochromocytoma-derived PC12 and an amperometric electrochemical detector (Antec, Leiden, The cell line (Greene and Tischler, 1976) which, in contrast to tissue Netherlands). The mobile phase consisted of 0.05 M citric acid, 0.05 M slices or primary neuronal cultures, consists of a homogeneous phosphoric acid, 1.5 mM octanesulfonic acid (sodium salt monohydrate, population of DA containing cells that have been demonstrated to Janssen Chimica, Beerse, Belgium), 0.2 mM EDTA, and 5% (v/v) meth- synthesize, store, and release DA in a similar manner as neurons anol, pH 2.5. The detector potential was set at ϩ750 mV versus a K/KCl reference electrode. Catecholamine content was calculated against stan- (Greene and Rein, 1977). Moreover, these cells are known to dards that had been subjected to the same treatment as the samples. contain both GSH and the enzyme systems involved in its metab- Cell survival assay. To determine cell survival, we used the 3-(4,5- olism (Pan and Perez-Polo, 1993; Sampath et al., 1994). dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay as described previously (Langeveld et al., 1992) with some adaptations. An MATERIALS AND METHODS MTT stock solution was prepared and stored at Ϫ20ЊC. On the day of the assay, 100 ␮l of MTT solution (5 mg/ml) was added to each well of the Materials. Radiolabeled DA and choline were obtained from Amersham culture dishes containing 1 ml of culture medium. Subsequently, incuba- International (Little Chalfont, UK). DL-␣-methyl-p-tyrosine (␣MPT) was tion of the cultures took place for 2 hr at 37ЊC under an atmosphere of from Research Biochemicals (Natick, MA), and reserpine was from De 5% CO2/95% air. Thereafter, the medium was removed and replaced by Onderlinge Pharmaceutische Groothandel (Utrecht, The Netherlands). Tis- 2 ml of dimethylsulfoxide (DMSO) to which 0.5% (v/v) FCS had been sue culture media and supplements were obtained from Gibco Netherlands added. After formazan solubilization by vibration on a plate shaker, a 150 BV (Breda, The Netherlands), whereas all other chemicals and drugs, unless ␮l sample was drawn from this solution and pipetted into a microtiter noted otherwise, were obtained from Sigma (St. Louis, MO). plate. The absorbance of each well was measured at 540 nm using the Cell culture. PC12 cells, originally described by Greene and Tischler 2 TitertekPlus microplate reader described above. Absorbance data were (1976), were grown in 80 cm tissue culture flasks containing DMEM/ calculated by subtracting the mean of “background” readings obtained Ham’s F-10 (1:1) supplemented with 15% fetal calf serum (FCS), 2 mM from identical incubations in the absence of cells. L-glutamine, penicillin (100 U/ml), and streptomycin (50 ␮g/ml) at 37ЊC Radioligand uptake. Cells were rinsed free of culture medium and under an atmosphere of 5% CO2/95% air. After 1 week, in which the preincubated for 30 min at 37ЊCin500␮l of a phosphate buffer, pH 7.3, culture medium had been changed once or twice, the medium was consisting of (in mM): NaCl 137, KCl 2.7, Na HPO 8, KH PO 1.5, removed and the cells were trypsinized according to Van Muiswinkel et 2 4 2 4 MgCl2 0.5, CaCl2 1.2, and D-(ϩ)-glucose 5. To this buffer 0.5% (w/v) al. (1995). After trypsinization, viable cells in the resulting cell suspension bovine serum albumin (BSA) had been added. Subsequently, the cultures were counted by trypan blue exclusion in a hemocytometer. Subsequently, were washed and incubated for 20 min with BSA containing buffer to cells were plated in 12-well culture dishes, containing 1 ml of culture 3 M 5 which [ H]DA (specific activity 45 Ci/mmol; final concentration 65 n ) medium in each well, at a density of 2.5 ϫ 10 cells/well for all experi- and/or [14C]choline (specific activity 55 mCi/mmol; final concentration 8.5 ments. Before culturing, both culture flasks and dishes had been pre- ␮M) had been added. Thereafter, cultures were washed seven times with coated with poly-L-lysine (100 ␮g/ml). Investigations were started when buffer before the accumulated radioactivity was extracted by a 30 min the cells had been in the culture dishes for at least 24 hr and the cultures incubation with acidified ethanol (95% ethanol/5% 0.1 N HCl) and were maintained, without a change of medium, for a maximum of 4 d quantified by liquid scintillation counting (Packard TriCarb 1900 AC depending on the type of experiment (see Results). liquid scintillation analyzer, Groningen, The Netherlands). If uptake GSH assay: sample preparation. Samples were prepared for the mea- inhibitors were used, they were added to the buffer at the start of the surement of cellular total GSH content, that is, reduced GSH ϩ oxidized preincubation period and remained present throughout the experiment. glutathione (GSSG), according to the method described by Redegeld et Protein content. Protein content was determined, according to the al. (1988), with the exception that 5-sulfosalicylic acid (SSA) was used method of Bradford (1976) and using BSA as a standard, in parallel instead of perchloric acid to extract GSH. Preparation of the samples was cultures to those used for the above described measurements. performed at 4ЊC. Cells were rinsed free of culture medium or phosphate Statistical analysis. Statistical analysis was performed using a two-tailed buffer (see below) before 250 ␮l of 2.5% (w/v) SSA was added. Subse- Student’s t test. quently, culture dishes were quickly frozen (Ϫ70ЊC) and thawed to ensure cell lysis. The lysate was centrifuged to remove denaturated protein. RESULTS From the supernatant a volume of 100 ␮l was taken to which 10 ␮lofH2O and, after 5 min, 20 ␮lofK3PO4 (3 M, pH 13.5) were added for incubation Effect of reserpine on GSH content at room temperature. After a 10 min period, the samples were then Under our culture conditions, ϳ95% of the total catecholamine neutralized with 50 ␮l of a 10% (w/v) SSA solution. In a subset of content of the PC12 cells consisted of DA (data not shown). As samples, instead of H2O, 10 ␮lof11mMN-ethyl-maleimide (NEM) was added to determine GSSG content. In these samples, alkaline hydrolysis expected, incubation of the cells for either 1 or 3 hr in a phosphate by K3PO4 was used for inactivation of excess NEM. buffer (see Materials and Methods) to which the granular uptake GSH assay: enzyme recycling procedure. GSH was determined spectro- inhibitor reserpine (50 nM; Carlsson, 1965) had been added led to photometrically according to Tietze (1969) with adaptations by Baker et a time-dependent decrease of DA content (Fig. 1A). However, al. (1990). The procedure was performed at room temperature. In short, 50 ␮l of GSH or GSSG standards, samples, or blanks was pipetted into whereas incubation of the PC12 cells in the buffer time- the wells of a microtiter plate. A buffer of 100 mM sodium phosphate and dependently lowered total GSH levels, this phenomenon was not 1mM EDTA (Fluka Chemie, Buchs, Switzerland; pH 7.5) was used to observed when reserpine was present (Fig. 1B). The decrease in prepare a reaction mixture consisting of 0.225 mM 5,5Ј-dithio-bis(2-nitro- GSH content under control condition during incubation with benzoic acid) (DTNB), 0.3 mM NADPH (tetrasodium salt, Boehringer Mannheim, Mannheim, Germany), and 2.8 U/ml GSSG reductase. Sub- buffer was prevented by addition of cystine (10–100 ␮M) which, sequently, 100 ␮l of freshly prepared reaction mixture was added to each after uptake, is reduced inside the cell into the GSH precursor of the wells. The absorbance at 405 nm (TitertekPlus Microplate Reader, cysteine. In fact,a3hrincubation with cystine induced a ICN Biomedicals, Amsterdam, The Netherlands) was measured immedi- concentration-dependent increase in GSH levels which, between ately after the addition of the reaction mixture and at intervals of 1 min experiments, averaged from 0.7–2.0 nmol/mg protein in the ab- thereafter during 6 min. Total GSH and GSSG content was calculated from the rate of increase in absorbance. sence of cystine to 8.7–18.3 nmol/mg protein in the presence of Cellular catecholamine content. Cells were rinsed free from culture 100 ␮M cystine (see legend to Fig. 2). In the presence of cystine, medium or phosphate buffer (see below) and subsequently lysed by a statistically significant elevation of GSH content was noted after adding 400 ␮l of ice-cold perchloric acid (0.2 N). After this procedure, the addition of reserpine (50 nM, Fig. 2), without alteration of its DA culture dishes were frozen and thawed and the lysate then centrifuged as described above. Aliquots of the supernatants were stored at Ϫ20ЊC content decreasing action (data not shown). Interestingly, this before the analysis of catecholamine content by HPLC with electrochem- effect of reserpine appeared to be inversely related to the con- ical detection. HPLC analysis was performed essentially as described centration of cystine in the buffer and thus, possibly, to cellular 6040 J. Neurosci., October 1, 1996, 16(19):6038–6045 Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine

Figure 1. Effect of reserpine on dopamine (DA) content (A) and glutathione (GSH) content (B) of PC12 cells. PC12 cells, cultured for 2 d, were rinsed free of culture medium and incubated for either 1 or 3 hr at 37ЊC in a phosphate buffer (see Materials and Methods) in the absence or presence of 50 nM reserpine, respectively. Subsequently, the cells were lysed and the DA, GSH, and protein content in parallel cultures was determined as detailed in Materials and Methods. Data represent the mean Ϯ SEM (n ϭ 5–6) from two independent experiments. Between experiments, at the start of incubation (i.e., t ϭ 0), DA content averaged from 1.8 to 3.3 nmol/mg protein, whereas GSH content averaged from 1.6 to 3.1 nmol/mg protein. Therefore, the data are expressed as percentage of DA or GSH content at t ϭ 0. *p Ͻ 0.001 versus control at t ϭ 0; **p Ͻ 0.001 versus respective time control.

GSH levels. In contrast to cystine, incubation of cells for 3 hr with experiments no extracellular GSH was detected in the incubation buffer containing 100 ␮ML-buthionine-S,R-sulfoximine (BSO), an buffer under either control or treatment conditions. inhibitor of GSH synthesis (Meister, 1991), led to a decrease of Treatment of PC12 cells with reserpine (50 nM) added to the GSH content below control values (see legend to Table 1). How- culture medium for up to 3 d induced a decrease of cellular DA ever, the presence of BSO did not prevent the reserpine (50 content down to ϳ1.5% of (time)control values (Table 2). The nM)-mediated effect on GSH levels from occurring (Table 1). In maximum level of DA depletion was already attained after a 24 hr all of the above described experiments, the GSSG levels remained incubation period and remained stable throughout the experi- below the limit of detection (0.05 nmol/well). Thus, we concluded ment. At the same time, in the presence of reserpine, the levels of that most of the GSH present in the cells was in the reduced form and decided to refrain from any further attempts at measurement Table 2. Effect of treatment of PC12 cells for up to 3 d with reserpine of GSSG in the follow-up experiments. Furthermore, in this set of on the GSH, DA, and DOPAC content

Content (% respective control) Table 1. Effect of BSO on the reserpine-induced increase in glutathione Treatment GSH DA DOPAC content in PC12 cells 24 hr no drug 100 Ϯ 12.3 100 Ϯ 0.3 100 Ϯ 0.8 a a b Glutathione content 24 hr reserpine (50 nM) 424 Ϯ 2.8 1.3 Ϯ 0.05 92.8 Ϯ 1.1 Treatment (% respective control) 48 hr no drug 100 Ϯ 13.8 100 Ϯ 1.5 100 Ϯ 5.3 a a No drug 100 Ϯ 3.4 48 hr reserpine (50 nM) 296 Ϯ 5.5 1.4 Ϯ 0.05 85.4 Ϯ 1.0 Reserpine (50 nM) 222 Ϯ 14.5* 72 hr no drug 100 Ϯ 8.3 100 Ϯ 0.9 100 Ϯ 1.5 a a c BSO (100 ␮M) 100 Ϯ 7.6 72 hr reserpine (50 nM) 414 Ϯ 1.7 1.4 Ϯ 0.02 88.0 Ϯ 1.2 BSO (100 ␮M) ϩ reserpine (50 nM) 276 Ϯ 9.2* After an initial 24 hr drug-free culture period, without a change of medium (see Materials and Methods), PC12 cells were kept in culture for an additional 24, 48, or After2dofculture, PC12 cells were rinsed free of medium and incubated for 3 hr 72 hr period in the absence (no drug) or presence of 50 nM reserpine, respectively. at 37ЊC in a phosphate buffer (see Materials and Methods) in the absence of drugs At the end of each incubation period, samples were prepared for determination of or in the presence of reserpine alone, BSO alone, or the combination of BSO and GSH, DA, and DOPAC content in parallel cultures, respectively, as detailed in reserpine, respectively. Data represent the mean Ϯ SEM (n ϭ 5–6) from two Materials and Methods. Subsequent to preparation, all samples were stored at independent experiments and are expressed, in case of reserpine alone, as percent- Ϫ20ЊC to ensure that determination of GSH, DA, and DOPAC content would take age of glutathione content after 3 hr in the absence of drugs and, in case of the place on the same day. Data represent the mean Ϯ SEM of a typical experiment combination of BSO and reserpine, as percentage of glutathione content after 3 hr performed in triplicate and are expressed as percentage of respective control (i.e., the in the presence of BSO alone. In the absence of drugs, the glutatione content after value obtained for the respective incubation period in the absence of drug). a 3 hr incubation amounted to 45.8 Ϯ 3.2% from control at t ϭ 0 hr, whereas in the a presence of BSO glutathione content aftera3hrincubation amounted to 26.7 Ϯ p Ͻ 0.001 vs respective control. 2.4% from control (mean Ϯ SEM; p Ͻ 0.001 vs incubation in absence of drugs). bp Ͻ 0.01 vs respective control. * p Ͻ 0.001 vs respective control. cp Ͻ 0.005 vs respective control. Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine J. Neurosci., October 1, 1996, 16(19):6038–6045 6041

mM) induced a concentration-dependent lowering of the DA level down to ϳ15% of control value (Fig. 3A), accompanied, however, by a decrease in DOPAC content below the detection limit at all drug concentrations used (data not shown). Concurrently, similar to the results obtained with reserpine, compared with control values total GSH content increased in a concentration-dependent manner in the presence of ␣MPT (Fig. 3B). After depletion of intracellular DA by treatment for 24 hr with 1 mM ␣MPT, incubation of the cells for 20 min in a phosphate buffer (see Materials and Methods) containing DA (0.1–1 mM) again increased the intracellular DA level while simultaneously decreasing GSH level in a concentration-dependent manner (Fig. 4). In contrast, PC12 cells that were not pretreated with ␣MPT but were incubated for 20 min with DA under similar experimental conditions showed only a marginal increase in intracellular DA content and no change in GSH level (data not shown). To control for possible “nonspecific,” non-DA-related drug effects on cellular GSH metabolism, human D384 glioma cells (Langeveld et al., 1992), not containing any detectable DA, were cultured in DMEM/F-10 containing 10% FCS and exposed to either reserpine or ␣MPT in the concentrations used for the experiments with PC12 cells. Incubation of the cells for up to 24 hr with either of the drugs did not induce any significant changes in cellular GSH content (data not shown). Furthermore, it is important to note that, under none of the above described experimental conditions in PC12 and D384 cells, neither reserpine nor ␣MPT affected cell survival as estimated by measurement of Figure 2. Effect of reserpine on glutathione (GSH) content of PC12 cells protein content (data not shown). in the presence of cystine. PC12 cells, cultured for 2 d, were rinsed free of culture medium and incubated for 3 hr in a phosphate buffer (see Mate- Effect of BSO on DA uptake rials and Methods) at 37ЊC with 10, 30, or 100 ␮M cystine in the absence To check whether radiolabeled DA and choline uptake in PC12 (control) or presence of 50 nM reserpine, respectively. Subsequently, the cells were lysed and the GSH and protein content in parallel cultures was cells, cultured under our experimental conditions, was mediated ϩ determined as detailed in Materials and Methods. Data represent the via a high-affinity membrane transport system, the Na concen- mean Ϯ SEM (n ϭ 9) from three independent experiments. Between tration in the incubation buffer was lowered by isomolar replace- experiments, the GSH content aftera3hrincubation in the absence of ment of NaCl by LiCl. Under these conditions, [3H]DA uptake reserpine averaged from 0.7 to 2.0 nmol/mg protein (no cystine), 1.5 to 4.4 was lowered to 18.2 Ϯ 0.7% of the control value [mean Ϯ SEM nmol/mg protein (10 ␮M cystine), 4.1 to 11.1 nmol/mg protein (30 ␮M 14 cystine), and 8.7 to 18.3 nmol/mg protein (100 ␮M cystine). Therefore, the (n ϭ 3); Student’s t test, p Ͻ 0.001], and [ C]choline uptake was data are expressed as percentage of respective GSH content at each lowered to 25 Ϯ 0.3% of control [mean Ϯ SEM (n ϭ 3); Student’s cystine concentration in the absence of reserpine. *p Ͻ 0.005 versus t test, p Ͻ 0.001]. Moreover, incubation of the cells with either the respective control. DA uptake inhibitor GBR 12909 (10 ␮M) (Andersen, 1989) or the NA uptake inhibitor desmethylimipramine (10 ␮M) (Iversen, the DA metabolite 3,4-dihydroxyphenylacetic acid (DOPAC) also 1973) reduced radiolabeled DA uptake down to 0.8 Ϯ 0.04 and decreased, albeit to a much smaller extent than those of DA, thus 1.0 Ϯ 0.03% of control, respectively [mean Ϯ SEM (n ϭ 3); demonstrating ongoing DA breakdown in the absence of granular Student’s t test, p Ͻ 0.001], whereas incubation with the choline storage (Table 2). Concomitant with these reserpine-induced ef- uptake inhibitor hemicholinium-3 (1 mM) (Birks and MacIntosh, fects on DA metabolism, a lasting and statistically significant 1957) reduced choline uptake to 26.6 Ϯ 0.8% of control [mean Ϯ increase in cellular total GSH content was observed after incuba- SEM (n ϭ 3); Student’s t test, p Ͻ 0.001]. Exposure of the PC12 tion with the drug for 1, 2, or 3 d (Table 2). Under these cells for 24 hr to 10 ␮M BSO added to the culture medium, circumstances, however, no effect of reserpine treatment on cell depleted the cellular GSH content down to 15.3 Ϯ 1.2% of control survival was detected, as measured with the MTT assay [0.65 Ϯ [mean Ϯ SEM (n ϭ 6); Student’s t test, p Ͻ 0.001]. Consistent 0.009 vs 0.68 Ϯ 0.009, 0.80 Ϯ 0.003 vs 0.77 Ϯ 0.02, 0.93 Ϯ 0.01 vs with previous data (Pan and Perez-Polo, 1993), in the presence of 0.97 Ϯ 0.02; mean Ϯ SEM (n ϭ 3) of absorbance in control versus this relatively low concentration of BSO no effect was observed on reserpine treatment group after 1, 2, or3dofincubation with cell survival (measured as protein content in treated and un- reserpine, respectively]. treated groups), which contrasted with the results obtained after a 24 hr incubation with higher concentrations of the drug, where Effect of ␣MPT on GSH content clear toxicity was induced without an additional effect on GSH To rule out the possibility that the reserpine-induced elevation of levels in the remaining cells (data not shown). After the incuba- 3 GSH levels was (partly) attributable to enhanced DA turnover tion with BSO (10 ␮M), the high-affinity uptake of [ H]DA was and not to inhibition of granular uptake and/or storage, we incu- decreased by ϳ25%, whereas no significant effect on [14C]choline bated the PC12 cells for 24 hr with the tyrosine hydroxlase (TH) uptake occurred (Fig. 5). At the same time, in BSO-treated inhibitor ␣-methyl-p-tyrosine (␣MPT) (Spector et al., 1965) added cultures a slight yet statistically significant lowering of the DA to the culture medium. Under these conditions, ␣MPT (0.1–1 content accompanied by an enhancement of DA turnover was 6042 J. Neurosci., October 1, 1996, 16(19):6038–6045 Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine

Figure 3. Effect of ␣-methyl-p-tyrosine (␣MPT) on dopamine (DA) content (A) and glutathione (GSH) content (B) of PC12 cells. After an initial 24 hr drug-free culture period, without a change of medium (see Materials and Methods), PC12 cells were kept in culture for an additional 24 hr period in the absence (0 ␮M) or presence of 100, 300, or 1000 ␮M ␣MPT, respectively. Subsequently, the cells were lysed and the DA, GSH, and protein content in parallel cultures was determined as detailed in Materials and Methods. Data represent the mean Ϯ SEM (n ϭ 5–6) from two independent experiments and are expressed as percentage of control (i.e., content after 48 hr of culture in the absence of ␣MPT). Under control conditions, between experiments the DA content averaged from 3.7 to 4.0 nmol/mg protein, whereas the GSH content averaged from 1.5 to 1.6 nmol/mg protein, respectively. *p Ͻ 0.001 versus control. noted [DA content: 94.5 Ϯ 1.0% vs control; DOPAC/DA ratio: its regulated release via exocytosis. Most information about the 108 Ϯ 1.5% vs control (mean Ϯ SEM, n ϭ 6; Student’s t test, granular transport mechanism(s) and storage mechanism(s) of p Ͻ 0.01)]. DA (and monoamines in general) has been obtained by using chromafﬁn granules isolated from adrenal medullary cells or PC12 DISCUSSION cells as experimental substrate (Roda et al., 1980). Thus, it is In general, DA, after its synthesis, is taken up and stored in presently known that the DAergic granular transport system con- specialized subcellular organelles, the storage granules, to ensure sists of at least two components: (1) a so-called vesicular mono-

Figure 4. Effect of exogenous dopamine (DA) on the DA content (A) and glutathione (GSH) content (B) of PC12 cells incubated in the presence of ␣-methyl-p-tyrosine (␣MPT). After an initial 24 hr drug-free culture period, without a change of medium (see Materials and Methods), PC12 cells were kept in culture for an additional 24 hr period in the absence or presence of 1 mM ␣MPT. Subsequently, the cells were rinsed free of culture medium and incubated for 20 min at 37ЊC in a phosphate buffer (see Materials and Methods) in the absence (0 ␮M) or presence of 100, 300, or 1000 ␮M DA, respectively. Thereafter, the cells were lysed and the DA, GSH, and protein content in parallel cultures was determined as detailed in Materials and Methods. Data represent the mean Ϯ SEM (n ϭ 6) from two independent experiments and are expressed as percentage of DA or GSH content in cells treated only with ␣MPT. In control cultures (i.e., cells not treated with any drug), the DA content amounted to 5.3 Ϯ 0.5 nmol/mg protein, whereas in cultures treated only with ␣MPT, the DA content amounted to 0.8 Ϯ 0.1 nmol/mg protein (mean Ϯ SEM). In control cultures, the GSH content amounted to 2.1 Ϯ 0.3 nmol/mg protein, whereas in cultures treated only with ␣MPT the GSH content amounted to 4.6 Ϯ 0.2 nmol/mg protein (mean Ϯ SEM). *p Ͻ 0.001 versus ␣MPT alone (i.e., 0 ␮M DA). Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine J. Neurosci., October 1, 1996, 16(19):6038–6045 6043

GSH being synthesized in the cell and suggested the existence of a (thus far unknown) distinct pool of cytoplasmic GSH intimately linked to the DA storage compartment of PC12 cells. This inter- pretation of our data is supported moreover by the fact that, irrespective of the amount of GSH present after incubation with increasing doses of cystine, at each cystine concentration the reserpine-mediated rise in GSH content, when calculated as ab- solute amounts increase, appeared to be very similar. Because it had been reported previously that reserpine affects the cellular GSH status by enhancing DA turnover (Spina and Cohen, 1989), we investigated, in addition, the effect of depletion of DA stores through inhibition of its synthesis. Although blockade of TH activity effectively lowered intracellular DA levels and, in contrast to reserpine, also reduced DA breakdown, again a significant increase in GSH content was detected that was of comparable magnitude to that observed with reserpine. Because inhibition of GSH formation did not prevent the observed effect, we concluded that the increased GSH level in the presence of reserpine was not attributable to enhanced synthesis but may be caused by a reduction in the use of GSH for cell function as a result of the emptying of the granular DA stores. This conclusion is supported moreover Figure 5. Effect of buthionine sulfoximine (BSO) on the uptake of by our data showing that, after an increase of GSH content caused [3H]dopamine (DA) and [14C]choline (Chol) in PC12 cells. After an initial by depletion of stored DA with ␣MPT, direct refilling of the DA drug-free culture period of 48 hr, without a change of medium (see stores by incubation with exogenous DA immediately reduced the Materials and Methods), PC12 cells were kept in culture for an additional period of 24 hr in the absence (control) or presence of 10 ␮M BSO, GSH level of the cells back to control values (i.e., GSH content of respectively. Subsequently, the culture medium was removed and radioli- cells not treated with ␣MPT). gand uptake was determined using a phosphate buffer to which both Although the described experiments demonstrated a relation- 3 14 [ H]dopamine and [ C]choline had been added, as detailed in Materials ship between the granular storage of DA and cellular GSH and Methods. Data represent the mean Ϯ SEM (n ϭ 6) from two independent experiments and are expressed as percentage of control (i.e., homeostasis, additional data were needed to establish whether uptake of either radioligand after 72 hr of culture in the absence of BSO). GSH itself plays a role in the granular transport of DA. In central Under control conditions, the uptake of [3H]dopamine and [14C]choline and peripheral neurons, released DA is recovered from the syn- amounted to 497,000 Ϯ 10,900 and 11,600 Ϯ 531 dpm (mean Ϯ SEM), apse via translocation into the cytoplasm by a Naϩ-dependent respectively. *p Ͻ 0.001 versus uptake of [3H]dopamine under control condition. membrane transporter (Rudnick and Clark, 1993). Subsequently, the DA taken up is protected from metabolic breakdown by storage in so-called presynaptic vesicles by the above described amine transporter (VMAT), structurally distinct from the plasma mechanism. Adrenal medulla-derived cells such as the PC12 cell membrane DA transporter, and (2) a vacuolar-type ATP-driven Hϩ pump, which provides the electrochemical gradient on which line are known to preferentially accumulate (radiolabeled) DA the transporter depends for its function (Johnson, 1988; Schuldi- from the external medium in the chromaffin granules in a way ner, 1994). Drugs such as reserpine and tetrabenazine deplete similar to that of DAergic neurons (Greene and Rein, 1977; intracellular DA stores by selectively interfering with transmitter Rudnick and Clark, 1993). Our results show that depletion of uptake via the VMAT. In the present set of experiments, we used GSH, by inhibition of its synthesis with the ␥-glutamylcysteine reserpine because it reportedly induces a long-lasting decrease in synthetase inhibitor BSO (Meister, 1991), significantly reduces the ϩ the DA level of PC12 cells at lower concentrations than tetraben- Na -dependent cellular uptake of radiolabeled DA in PC12 cells. azine and, moreover, in contrast to tetrabenazine, specifically In addition to DA (and NA), PC12 cells are known to transport blocks the transmitter (DA) recognition site on the VMAT (radiolabeled) choline, the precursor of ACh, across the plasma (Erickson and Eiden, 1993; Schuldiner et al., 1993). membrane using a transporter molecule (Melega and Howard, Our data show that, as reported previously (Greene and Rein, 1981). In our experiments, depletion of GSH had no effect on the 1977), inhibition of DA transport across the storage granule uptake of radiolabeled choline. Thus, although our experimental membrane in PC12 cells by a maximally effective concentration of setup does not allow us to differentiate between the transport of reserpine (our unpublished observations) induces a fast, almost DA across the plasma membrane or the chromaffin granule mem- total, and long-lasting depletion of intracellular DA without af- brane, in our opinion, by inference, it is more likely that the fecting cell survival. This demonstrated that, in PC12 cells cul- diminished uptake of DA reflects a lesion in the chromaffin tured under our conditions the major part of intracellularly syn- granule membrane transport system rather than in its plasma thesized, endogenous DA is sequestered in the storage granules membrane counterpart. Contrary to expectation, however, the and does not exert any toxic effect on the cells, as is observed in supposed defect in DA storage led only to a small, although case of exogenous DA (Michel and Hefti, 1990). Interestingly, the statistically significant, rise in the indices of DA metabolism, i.e., reserpine-mediated emptying of the granular DA stores was ac- a decrease in the steady-state level of endogenous DA in combi- companied by a substantial and sustained rise in the intracellular nation with an increase in DA turnover. Although further inves- total GSH (i.e., reduced GSH ϩ GSSG) content. Using the GSH tigation of this issue is needed, the apparent contradiction may be precursor cystine to stimulate GSH formation, the extent of this explained by the fact that, under our experimental conditions, the reserpine-induced effect appeared to depend on the amount of depletion of GSH over a 24 hr time period caused only a relatively 6044 J. Neurosci., October 1, 1996, 16(19):6038–6045 Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine small decrease in DA accumulation of which the effect on endog- pump, eventually leads to nigro-striatal cell death through massive enous DA metabolism is still unknown. oxidative stress. In this context, it is of interest to note that Interestingly, the above described data are very similar to those oxidative degeneration of DAergic neurons indeed has been ob- obtained previously by Elroy-Stein and Groner (1988), who found served after interference with the vesicular electrochemical gra- a 50–80% reduction in the uptake of monoamines, including DA, dient (Cubells et al., 1994). but no change in the uptake of choline in transformed PC12 cells overexpressing the human Cu/Zn-superoxide dismutase gene. Us- REFERENCES ing elegant techniques, they demonstrated that the impaired monoamine accumulation was caused by a diminished electro- Adams JD, Odunze IN (1991) Oxygen free radicals and Parkinson’s disease. Free Radic Biol Med 10:161–169. chemical gradient across the chromaffin granule membrane in the Andersen PH (1989) The dopamine uptake inhibitor GBR 12909: selec- transformed clones. As mentioned already, the electrochemical tivity and molecular mechanism of action. Eur J Pharmacol gradient is the driving force for the transport of monoamines, such 166:493–504. as DA, into the storage granules and is generated by a vacuolar- Baker MA, Cerniglia GJ, Zaman A (1990) Microtiter plate assay for the ϩ measurement of glutathione and glutathione disulfide in large numbers type, ATP-dependent H pump. This subclass of proton pumps, of biological samples. Anal Biochem 190:360–365. including the one present in the chromaffin granule membrane, is Birks RI, MacIntosh FC (1957) Acetylcholine metabolism at nerve- known to be extremely sensitive to sulfhydryl modification by endings. Br Med Bull 13:157–161. Bindoli A, Rigobello MP, Deeble DJ (1992) Biochemical and toxicolog- oxidants such as H2O2 (Rudnick, 1986), against which it is protected by reductants like GSH (Dschida and Bowman, 1995). ical properties of the oxidation products of catecholamines. Free Radic Biol Med 13:391–405. Thus, malfunction of the pump may be attributable either to Bradford M (1976) A rapid and sensitive method for the quantification increased production of ROS, such as occurs in cells with a high of microgram quantities of protein utilizing the principle of protein-dye superoxide dismutase activity (Elroy-Stein and Groner, 1988; binding. Anal Biochem 72:248–254. Schickler et al., 1989), or to decreased protection, for instance, Carlsson A (1965) Drugs which block the storage of 5-hydroxytryptamine and related amines. Handb Exp Pharmacol 19:529–592. after reduction of cellular GSH levels. Moreover, it is conceivable Cubells JF, Rayport S, Rajendran G, Sulzer D (1994) Metamphetamine that oxidative damage to the proton pump also forms the basis of neurotoxicity involves vacuolation of endocytic organelles and our previous observations demonstrating a preferential inhibition dopamine-dependent intracellular oxidative stress. J Neurosci of the vesicular storage of DA in the nigro-striatal system of the 14:2260–2271. rat brain by H O (Langeveld et al., 1995). At present, experi- Dexter DT, Sian J, Rose S, Hindmarsh JG, Mann VM, Cooper JM, Wells 2 2 FR, Daniel SE, Lees AJ, Schapira AHV, Jenner P, Marsden CD (1994) ments along the lines described by Elroy-Stein and Groner (1988) Indices of oxidative stress and mitochondrial function in individuals are being implemented in our laboratory to identify, possibly, a with incidental Lewy body disease. Ann Neurol 35:38–44. proton pump deficit as the cause of the diminished uptake of DA DiMonte DA, Chan P, Sandy MS (1992) Glutathione in Parkinson’s in GSH depleted PC12 cells. disease: a link between oxidative stress and mitochondrial damage? Ann In conclusion, our data, taken together, strongly suggest that Neurol 32:S111–S115. Dschida WJA, Bowman BJ (1995) The vacuolar ATPase: sulfite stabili- GSH is involved in the uptake of DA in the chromaffin granules zation and the mechanism of nitrate inactivation. J Biol Chem of PC12 cells. Although further research is clearly warranted and 270:1557–1563. alternative explanations are available, considering the fact that Elroy-Stein O, Groner Y (1988) Impaired neurotransmitter uptake in DA is easily oxidized (Bindoli et al., 1992), it is tempting to PC12 cells overexpressing human Cu/Zn-superoxide dismutase: implication for gene dosage effects in Down’s syndrome. Cell 52:259–267. propose a mechanism in which GSH is used to protect susceptible Erickson JD, Eiden LE (1993) Functional identification and molecular parts of the transport system against the potentially toxic interac- cloning of a human brain vesicle monoamine transporter. J Neurochem tion with DA oxidation products during DA uptake. Apparently, 61:2314–2317. during this process GSH is transformed into a compound, for Gibb WRG, Lees AJ (1991) Anatomy, pigmentation, ventral and dorsal instance, a glutathionyl conjugate (Nappi and Vass, 1994), which subpopulations of the substantia nigra, and differential cell death in Parkinson’s disease. J Neurol Neurosurg Psych 54:388–396. escapes detection by our enzyme recycling method, which is rather Greene LA, Rein G (1977) Release, storage and uptake of cat- specific for GSH and GSSG (Redegeld et al., 1988). Thus, a echolamines by a clonal cell line of nerve growth factor (NGF) respon- reduction in granular storage of DA, in our experiments induced sive pheochromocytoma cells. Brain Res 129:247–263. by reserpine or ␣MPT, may lead to an increase in measurable Greene LA, Tischler AS (1976) Establishment of a noradrenergic clonal GSH content. Conversely, a lowering of GSH content will leave line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. Proc Natl Acad Sci USA 73:2424–2428. the granular DA transport system relatively unprotected against Hirsch EC, Brandel JP, Galle P, Javoy-Agid F (1991) Iron and aluminum oxidative damage, resulting in a diminished storage of the increase in the substantia nigra of patients with Parkinson’s disease: an transmitter. X-ray microanalysis. J Neurochem 56:446–451. Notwithstanding the fact that differences in molecular structure Iversen LL (1973) Catecholamine uptake processes. Br Med Bull 29:130–135. and function of VMATs have been noted (Peter et al., 1994) and Jenner P (1993) Altered mitochondrial function, iron metabolism and that, in general, data obtained in adrenal medulla-derived PC12 glutathione levels in Parkinson’s disease. Acta Neurol Scand 87[Suppl cells are not directly extrapolatable to DAergic neurons, concen- 146]:6–13. tration of DA in synaptic vesicles of neurons is thought to involve Johnson Jr RG (1988) Accumulation of biological amines into chromaf- similar mechanisms as those active in chromaffin granules (Schul- fin granules: a model for hormone and neurotransmitter transport. Physiol Rev 68:233–307. diner, 1994). GSH is present in neurons (Pileblad et al., 1991), and Langeveld CH, Van Waas MP, Stoof JC, Sutanto W, De Kloet ER, in PD loss of GSH has been implicated as an early event in the Wolbers JG, Heimans JJ (1992) Implication of glucocorticoid recep- pathogenesis of the disorder (Dexter et al., 1994). Looking at our tors in the stimulation of human glioma cell proliferation by dexameth- results, it seems plausible to suggest that the consequent shift in asone. J Neurosci Res 31:524–531. Langeveld CH, Schepens E, Stoof JC, Bast A, Drukarch B (1995) Dif- the neuronal redox balance will initiate the self-sustaining cycle ferential sensitivity to hydrogen peroxide of dopaminergic and norad- already described in the introductory remarks, in which a dysfunc- renergic neurotransmission in rat brain slices. Free Radic Biol Med tion in vesicular storage of DA, probably at the level of the proton 19:209–217. Drukarch et al. • Role of Glutathione in Granular Storage of Dopamine J. Neurosci., October 1, 1996, 16(19):6038–6045 6045

Maker HS, Weiss C, Silides DJ, Cohen G (1981) Coupling of dopamine storage of biogenic amine neurotransmitters. Biochim Biophys Acta oxidation (monoamine oxidase activity) to glutathione oxidation via the 1144:249–263. generation of hydrogen peroxide in rat brain homogenates. J Neuro- Sampath D, Jackson GR, Werrbach-Perez K, Perez-Polo JR (1994) Ef- chem 36:589–593. fects of nerve growth factor on glutathione peroxidase and catalase in Meister A (1991) Glutathione deficiency produced by inhibition of its PC12 cells. J Neurochem 62:2476–2479. synthesis, and its reversal: applications in research and therapy. Phar- Schickler M, Knobler H, Avraham KB, Elroy-Stein O, Groner Y (1989) macol Ther 51:155–194. Diminished serotonin uptake in platelets of transgenic mice with in- Meister A, Andersson ME (1983) Glutathione. Annu Rev Biochem creased Cu/Zn-superoxide dismutase activity. EMBO J 8:1385–1392. 52:711–760. Schuldiner S (1994) A molecular glimpse of vesicular monoamine trans- Melega WP, Howard BD (1981) Choline and acetylcholine metabolism porters. J Neurochem 62:2067–2078. in PC12 secretory cells. Biochemistry 20:4477–4483. Schuldiner S, Liu Y, Edwards RH (1993) Reserpine binding to a vesic- Michel PP, Hefti F (1990) Toxicity of 6-hydroxydopamine and dopamine ular amine transporter expressed in Chinese hamster ovary fibroblasts. for dopaminergic neurons in culture. J Neurosci Res 26:428–435. J Biol Chem 268:29–34. Nappi AJ, Vass E (1994) The effects of glutathione and ascorbic acid on Sofic E, Lange KW, Jellinger K, Riederer P (1992) Reduced and oxidized the oxidations of 6-hydroxydopa and 6-hydroxydopamine. Biochim Bio- glutathione in the substantia nigra of patients with Parkinson’s disease. phys Acta 1201:498–504. Neurosci Lett 142:128–130. Olanow CW (1992) An introduction to the free radical hypothesis in Spector S, Sjoerdsma A, Udenfriend S (1965) Blockade of endogenous Parkinson’s disease. Ann Neurol 32:S2–S9. norepinephrine synthesis by alpha-methyl-tyrosine, an inhibitor of ty- Pan Z, Perez-Polo JR (1993) Role of nerve growth factor in oxidant rosine hydroxylase. J Pharmacol Exp Ther 147:86–95. homeostasis: glutathione metabolism. J Neurochem 61:1713–1721. Spina MB, Cohen G (1988) Exposure of school synaptosomes to L-dopa Peter D, Jimenez J, Liu Y, Kim J, Edwards RH (1994) The chromaffin increases levels of oxidized glutathione. J Pharmacol Exp Ther granule and synaptic vesicle amine transporters differ in substrate recognition and sensitivity to inhibitors. J Biol Chem 269:7231–7237. 247:502–507. Pileblad E, Eriksson PS, Hansson E (1991) The presence of glutathione Spina MB, Cohen G (1989) Dopamine turnover and glutathione oxida- in primary neuronal and astroglial cultures from rat cerebral cortex and tion: implications for Parkinson’s disease. Proc Natl Acad Sci USA brain stem. J Neural Transm 86:43–49. 86:1398–1400. Redegeld FA, Van Opstal MA, Houdkamp E, Van Bennekom WP (1988) Tietze F (1969) Enzymatic method for quantification of total and oxi- Determination of glutathione in biological material by flow-injection dized glutathione: applications to mammalian blood and other tissue. analysis using an enzymatic recycling reaction. Anal Biochem Anal Biochem 27:502–522. 174:489–495. Van Muiswinkel FL, Jongenelen CAM, Schepens HTWJ, Stoof JC, Riederer P, Sofic E, Rausch W-D, Schmidt B, Reynolds GP, Jellinger K, Drukarch B (1995) Effects of chronic activation of dopamine D-2 Youdim MBH (1989) Transition metals, ferritin, glutathione, and receptors in cultures of rat fetal dopaminergic neurons: indications for ascorbic acid in Parkinsonian brains. J Neurochem 52:515–520. alterations in functional activity. Dev Brain Res 85:128–136. Roda LG, Nolan JA, Seung UK, Hogue-Angeletti RA (1980) Isolation Westerink BHC, Mulder TBA (1981) Determination of picomole and characterization of chromaffin granules from a pheochromocytoma amounts of dopamine, noradrenaline, 3,4-dihydroxyphenylalanine, (PC 12) cell line. Exp Cell Res 128:103–109. 3,4-dihydroxyphenylacetic acid, homovanillic acid, and 5- Rudnick G (1986) ATP-driven Hϩ pumping into intracellular organelles. hydroxyindoleacetic acid in nervous tissue after one-step purification on Annu Rev Physiol 48:403–413. Sephadex G-10, using high-performance liquid chromatography with a Rudnick G, Clark J (1993) From synapse to vesicle: the reuptake and novel type of electrochemical detection. J Neurochem 36:1449–1462.